RAID is not backup

Simply running RAID on your desktop system is not backup.

A backup is a separate copy, stored on a separate system, and ideally you have another copy offsite. You can follow the line of disaster: If your main system explodes, you have a backup on your spare system; if that system explodes, you have a copy on your main system; but if your home explodes, you have a copy offsite. Enough exploding.

What RAID is great for is redundancy, and as we all know, redundant systems are there to increase system reliability through duplication of functions.

In the case of RAID 1, 5, and 6, an entire drive can fail, and the parity (or mirror) the system uses can enable it to recover the local data. RAID 5 can handle a single drive failure from three or more drives, while RAID 6 can handle two drives failing running with four drives or more — this was added because the time to rebuild large arrays inevitably increases the window of opportunity of a second drive failing.

What RAID doesn’t do is create a backup, so if the entire RAID fails — and it does happen — you lose all your files, and recovering data from a RAID is a huge pain. It also doesn’t necessarily protect against bitrot — the undetected deterioration of digital data on a storage medium. Advanced filesystems, such as ZFS and Btfs, offer checksum protection, but these aren’t a realistic option for Windows users. So we’re back to suggesting you need a real backup to secure data safely.